All of lore.kernel.org
 help / color / mirror / Atom feed
* Aacraid - terrible performance on RHES3 pedge 2450
@ 2004-10-25 10:31 Zoe May
  2004-10-25 12:22 ` Matt Domsch
  0 siblings, 1 reply; 6+ messages in thread
From: Zoe May @ 2004-10-25 10:31 UTC (permalink / raw)
  To: linux-raid


I am experiencing disk performance issues (very slow) with hardware and
software combo listed below.

Poweredge 2450 
Perc 3/si controller.
RH linux ES3 (2.4.21-4.Elsmp)
4 x 73GB scsi (In a RAID 5 set)

This server was previously running RH 7.3 with a 34GB mirror set with no
issue

So far I have updated the firmware of the card to Version 2.8.0.6092,
Rev. A13
And have updated the aacraid driver to aacraid-1.1.4-2302
Neither of the above have had any impact.

This machine is freshly installed and has no load to speak of yet
As an indicator of performance here is what I get from hdparm.


/dev/sda2:
 Timing buffer-cache reads:   780 MB in  2.00 seconds = 390.00 MB/sec
 Timing buffered disk reads:   22 MB in  3.20 seconds =   6.88 MB/sec
[root@benidorm root]# hdparm -tT /dev/sda1

/dev/sda1:
 Timing buffer-cache reads:   788 MB in  2.00 seconds = 394.00 MB/sec
 Timing buffered disk reads:   26 MB in  3.02 seconds =   8.61 MB/sec

When I do disk writes the system pretty much locks up in terms of doing
anything else until it's finished.

The following is a vmstat when the system is idle

vmstat 1
procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy
wa id
 0  0      0 1541840  23028 127984    0    0    22     4   80    85  1
0  1 98
 0  0      0 1541840  23028 127984    0    0     0     0  166   165  0
0  0 100
 0  0      0 1541840  23028 127984    0    0     0     0  171   172  0
0  0 100
 0  0      0 1541840  23028 127984    0    0     0     0  170   179  0
0  0 100
 0  0      0 1541824  23044 127984    0    0     0    48  174   180  0
0  0 100
 0  0      0 1541824  23044 127984    0    0     0     0  166   165  0
0  0 100
 0  0      0 1541824  23044 127984    0    0     0     0  171   160  0
0  0 100
 0  0      0 1541824  23044 127984    0    0     0     0  164   170  0
0  0 100
 2  0      0 1541824  23044 127984    0    0     0     0  169   167  0
0  0 100
 0  0      0 1541816  23052 127984    0    0     0    44  178   178  0
0  0 100
 0  0      0 1541816  23052 127984    0    0     0     0  163   161  0
0  0 100
 0  0      0 1541816  23052 127984    0    0     0     0  164   160  0
0  0 100
 1  0      0 1541816  23052 127984    0    0     0     0  163   165  0
0  0 100

This is what happens when I run 'dd if=/dev/zero of=test bs=1024000
count=500'


[root@benidorm raid]# vmstat 1
procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy
wa id
 0  0      0 1539248  23168 127984    0    0    22     4   80    86  1
0  1 98
 0  0      0 1539248  23168 127984    0    0     0     0  202   226  0
0  0 100
 0  0      0 1539248  23168 127984    0    0     0     0  445   458  0
0  0 99
 0  1      0 1539196  23196 127984    0    0    28     0  239   258  0
0  5 95
 2  0      0 1438304  23356 227984    0    0    28 14992  561   329  0
46 11 43
 1  0      0 1438304  23356 227984    0    0     0     0  254   238  0
0  0 100
 2  0      0 1438304  23356 227984    0    0     0     0  348   331  0
0  0 100
 0  0      0 1438304  23356 227984    0    0     0     0  311   404  0
0  0 100
 0  0      0 1438304  23356 227984    0    0     0     0  213   218  0
0  0 100
 1  2      0 1437672  23360 227984    0    0     0 63704  826   371  0
4 57 38
 0  2      0 1437668  23360 227984    0    0     0  8192  313   184  0
0 50 50
 0  2      0 1437664  23360 227984    0    0     0  8192  310   166  0
0 50 50
 0  2      0 1437652  23360 227984    0    0     0  5040  337   199  0
1 49 50
 2  2      0 1437652  23360 227984    0    0     0     0  332   190  0
0 50 50
 0  2      0 1437652  23360 227984    0    0     0   370  522   379  0
0 50 50
 0  2      0 1437652  23360 227984    0    0     0     0  417   285  0
0 50 50
 0  0      0 1437644  23372 227984    0    0     0   120  322   216  0
0  7 92
 0  0      0 1437644  23372 227984    0    0     0     0  275   254  1
0  0 100
procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy
wa id
 3  0      0 1437644  23372 227984    0    0     0     0  279   282  0
0  0 100
 0  0      0 1437636  23380 227984    0    0     0    16  597   617  1
0 15 83
 0  0      0 1437632  23380 227984    0    0     0     0  603   598  3
0  0 97
 0  0      0 1437632  23380 227984    0    0     0     0  192   188  0
0  0 100
 0  0      0 1437632  23380 227984    0    0     0     0  200   200  0
1  0 99
 1  0      0 1437632  23380 227984    0    0     0     0  188   189  0
0  0 100
 1  0      0 1437628  23388 227984    0    0     0    16  215   204  0
0  0 100
 0  0      0 1437628  23388 227984    0    0     0     0  289   296  1
0  0 99

[root@benidorm raid]# vmstat 1
procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy
wa id
 2  0      0 1225320  23700 431252    0    0    22    12   81    86  1
0  1 98
 1  2      0 1113708  23816 538284    0    0     4 42376  334   176  0
51  8 42
 1  3      0 1053544  23876 595980    0    0     4 23108  607   289  0
29 72  0
 0  2      0 1020184  23908 627984    0    0     0 10492  541   388  0
14 50 36
 0  2      0 1020184  23908 627984    0    0     0  8192  457   319  1
0 50 49
 0  2      0 1020184  23908 627984    0    0     0  8192  374   244  2
0 50 47
 0  2      0 1020184  23908 627984    0    0     0 10560  411   284  0
0 50 50
 1  2      0 1020184  23908 627984    0    0     0  6720  307   178  2
0 50 47
 0  2      0 1020184  23908 627984    0    0     0  8256  398   333  1
0 50 48
 0  2      0 1020184  23908 627984    0    0     0 10240  305   174  1
0 50 48
 0  2      0 1020184  23908 627984    0    0     0  8256  371   233  1
1 49 49
 0  2      0 1020184  23908 627984    0    0     0 10368  319   182  2
1 48 48
 0  2      0 1020184  23908 627984    0    0     0 10432  313   168  5
1 49 45
 0  2      0 1020184  23908 627984    0    0     0  8640  294   171  1
3 52 44
 0  2      0 1020184  23908 627984    0    0     0  8384  297   171  0
0 49 50
 0  2      0 1020184  23908 627984    0    0     0  8256  309   178  0
1 49 50
 0  2      0 1020184  23908 627984    0    0     0  8512  297   179  0
1 49 50
 0  2      0 1020184  23908 627984    0    0     0  8832  369   213  0
1 49 50
procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy
wa id
 1  2      0 1020184  23908 627984    0    0     0 10240  361   238  1
0 49 50
 0  2      0 1020184  23908 627984    0    0     0  8128  300   181  0
0 50 50
 0  2      0 1020184  23908 627984    0    0     0  8192  332   225  0
0 51 50
 0  2      0 1020184  23908 627984    0    0     0 10624  488   343  1
1 48 50
 0  2      0 1020184  23908 627984    0    0     0  8320  304   167  0
1 48 50
 0  2      0 1020184  23908 627984    0    0     0  8704  302   180  1
1 49 50
 0  2      0 1020184  23908 627984    0    0     0 10304  308   178  0
1 49 50
 0  2      0 1020184  23908 627984    0    0     0  8384  304   178  1
0 48 50
 0  2      0 1020184  23908 627984    0    0     0  8960  306   196  0
2 48 50
 0  2      0 1020184  23908 627984    0    0     0  8192  316   191  1
0 49 50
 0  2      0 1020184  23908 627984    0    0     0  8192  348   239  1
1 48 50
 0  2      0 1020184  23908 627984    0    0     0  8192  301   176  1
1 48 50
 0  2      0 1020184  23908 627984    0    0     0 10240  390   263  1
1 48 50
 0  3      0 1020184  23908 627984    0    0     0  8188  399   271  2
2 47 50
 0  3      0 1020184  23908 627984    0    0     0 11136  331   185  2
0 48 50
 1  3      0 1020184  23908 627984    0    0     0  8256  353   233  1
1 47 50
 0  3      0 1020184  23908 627984    0    0     0  8192  302   182  2
0 54 43
 0  3      0 1020180  23908 627984    0    0     0  7132  292   186  3
2 95  0
procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy
wa id
 1  4      0 1019896  23908 627988    0    0     0  8132  350   217  3
1 96  0
 0  4      0 1019896  23908 627988    0    0     0  7380  299   183  2
1 96  0
 1  4      0 1019896  23908 627988    0    0     0  7292  361   223  2
3 96  0
 1  4      0 1019896  23908 627988    0    0     0 10920  357   207  1
3 96  0
 0  4      0 1019892  23916 627988    0    0     0  9916  327   192  2
8 90  0
 1  4      0 1019892  23916 627988    0    0     0  8832  333   180  5
3 92  0
 0  4      0 1019892  23916 627988    0    0     0 10560  327   178  0
0 100  0
 0  4      0 1019880  23924 627988    0    0     0 10892  326   175  0
1 99  0
 0  4      0 1019880  23924 627988    0    0     0  8384  294   173  0
0 100  0
 1  4      0 1019880  23924 627988    0    0     0  8832  313   168  0
1 99  0
 1  4      0 1019880  23924 627988    0    0     0 10368  432   332  0
0 100  0
 0  4      0 1019880  23924 627988    0    0     0  8128  371   290  0
0 100  0
 0  4      0 1019880  23924 627988    0    0     0  8192  307   166  0
1 100  0
 0  4      0 1019880  23924 627988    0    0     0  8576  297   168  0
0 100  0
 0  4      0 1019880  23924 627988    0    0     0  8768  299   175  0
0 100  0
 1  3      0 1019880  23924 627988    0    0     0  5944  299   177  0
0 100  0
 1  3      0 1019880  23924 627988    0    0     0     0  306   156  0
0 100  0
 0  3      0 1019880  23924 627988    0    0     0     0  300   158  0
0 100  0



This is the container config output from aafcli

_container show > cache 0
Executing: container show cache 0

Global Container Read Cache Size  : 0
Global Container Write Cache Size : 51380224

Read Cache Setting        : ENABLE
Write Cache Setting       : ENABLE ALWAYS
Write Cache Status        : Active, not protected, battery not present

The write cache setting was previously 'enable when protected'. I
changed it to enable always but it had no impact.
I'm not sure where to go with this now. Possibly trying to boot off
knoppix or something to see if a different kernel changes things.
If anyone has any suggestions I'd be very grateful.
Thanks
Zoe 


^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Aacraid - terrible performance on RHES3 pedge 2450
  2004-10-25 10:31 Zoe May
@ 2004-10-25 12:22 ` Matt Domsch
  0 siblings, 0 replies; 6+ messages in thread
From: Matt Domsch @ 2004-10-25 12:22 UTC (permalink / raw)
  To: Zoe May; +Cc: linux-raid

On Mon, Oct 25, 2004 at 11:31:07AM +0100, Zoe May wrote:
> 
> I am experiencing disk performance issues (very slow) with hardware and
> software combo listed below.
> 
> Poweredge 2450 
> Perc 3/si controller.
> RH linux ES3 (2.4.21-4.Elsmp)

up2date your system to kernel-smp-2.4.21-20.EL and see how much that
helps.

You're not the only person who has seen slow disk performance with the
-4.EL kernel.

Thanks,
Matt

-- 
Matt Domsch
Sr. Software Engineer, Lead Engineer
Dell Linux Solutions linux.dell.com & www.dell.com/linux
Linux on Dell mailing lists @ http://lists.us.dell.com

^ permalink raw reply	[flat|nested] 6+ messages in thread

* RE: Aacraid - terrible performance on RHES3 pedge 2450
@ 2004-10-25 12:28 Salyzyn, Mark
  2004-10-25 13:01 ` Matt Domsch
  0 siblings, 1 reply; 6+ messages in thread
From: Salyzyn, Mark @ 2004-10-25 12:28 UTC (permalink / raw)
  To: Matt Domsch, Zoe May; +Cc: linux-raid

Vary_io may be the reason for this.

The driver needs to have .vary_io set to 1.

There are no kernel clues to its existence (that I know of), but the
driver has an SCSI_HAS_VARY_IO to enable. Our build system will produce
correct binaries, but the DKMS tool does not differentiate.

There is a need to ship a different driver source in the DKMS package
for RHEL3 as a result (Add a #define SCSI_HAS_VARY_IO in the aacraid.h
file).

Sincerely -- Mark Salyzyn

-----Original Message-----
From: linux-raid-owner@vger.kernel.org
[mailto:linux-raid-owner@vger.kernel.org] On Behalf Of Matt Domsch
Sent: Monday, October 25, 2004 8:22 AM
To: Zoe May
Cc: linux-raid@vger.kernel.org
Subject: Re: Aacraid - terrible performance on RHES3 pedge 2450

On Mon, Oct 25, 2004 at 11:31:07AM +0100, Zoe May wrote:
> 
> I am experiencing disk performance issues (very slow) with hardware
and
> software combo listed below.
> 
> Poweredge 2450 
> Perc 3/si controller.
> RH linux ES3 (2.4.21-4.Elsmp)

up2date your system to kernel-smp-2.4.21-20.EL and see how much that
helps.

You're not the only person who has seen slow disk performance with the
-4.EL kernel.

Thanks,
Matt

-- 
Matt Domsch
Sr. Software Engineer, Lead Engineer
Dell Linux Solutions linux.dell.com & www.dell.com/linux
Linux on Dell mailing lists @ http://lists.us.dell.com
-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Aacraid - terrible performance on RHES3 pedge 2450
  2004-10-25 12:28 Aacraid - terrible performance on RHES3 pedge 2450 Salyzyn, Mark
@ 2004-10-25 13:01 ` Matt Domsch
  0 siblings, 0 replies; 6+ messages in thread
From: Matt Domsch @ 2004-10-25 13:01 UTC (permalink / raw)
  To: Salyzyn, Mark; +Cc: Zoe May, linux-raid

On Mon, Oct 25, 2004 at 08:28:47AM -0400, Salyzyn, Mark wrote:
> Vary_io may be the reason for this.
> 
> The driver needs to have .vary_io set to 1.
> 
> There are no kernel clues to its existence (that I know of), but the
> driver has an SCSI_HAS_VARY_IO to enable. Our build system will produce
> correct binaries, but the DKMS tool does not differentiate.
> 
> There is a need to ship a different driver source in the DKMS package
> for RHEL3 as a result (Add a #define SCSI_HAS_VARY_IO in the aacraid.h
> file).

We could do this easily in DKMS with its PATCH_MATCH mechanism, making
it match RHEL kernel names and apply the patch only then which sets
.vary_io=1.

-- 
Matt Domsch
Sr. Software Engineer, Lead Engineer
Dell Linux Solutions linux.dell.com & www.dell.com/linux
Linux on Dell mailing lists @ http://lists.us.dell.com

^ permalink raw reply	[flat|nested] 6+ messages in thread

* RE: Aacraid - terrible performance on RHES3 pedge 2450
@ 2004-11-01 15:51 Zoe May
  0 siblings, 0 replies; 6+ messages in thread
From: Zoe May @ 2004-11-01 15:51 UTC (permalink / raw)
  To: Matt Domsch, Salyzyn, Mark; +Cc: linux-raid

Thanks for the info

I've upgraded the kernel to 2.4.21-20 and have a very slight increase in
performance but unfortunately not much to write home about.
I have limited experience with kernel compilation and I'm a little
confused as to which aacraid.h I would have to add the #define
SCSI_HAS_VARY_IO 1 to.
 
I have the following aacraid.h files

/usr/src/aacraid-1.1.4-2302/aacraid.h
/usr/src/linux-2.4.21-20.EL/drivers/addon/aacraid_10102/aacraid.h
/usr/src/linux-2.4.21-20.EL/drivers/scsi/aacraid/aacraid.h

Ordinarily I would modify the .h file in the scsi/aacraid dir and then
go through the full kernel recompile but I have updated aacraid using
the Dell rpm. I assume the version in this folder is the old version
that shipped with the 2.4.21-20 source. If there is a way that I can
recompile just the aacraid module without doing a full kernel recompile
I would prefer to do this .


Zoe


-----Original Message-----
From: Matt Domsch [mailto:Matt_Domsch@dell.com] 
Sent: Monday, October 25, 2004 2:02 PM
To: Salyzyn, Mark
Cc: Zoe May; linux-raid@vger.kernel.org
Subject: Re: Aacraid - terrible performance on RHES3 pedge 2450

On Mon, Oct 25, 2004 at 08:28:47AM -0400, Salyzyn, Mark wrote:
> Vary_io may be the reason for this.
> 
> The driver needs to have .vary_io set to 1.
> 
> There are no kernel clues to its existence (that I know of), but the 
> driver has an SCSI_HAS_VARY_IO to enable. Our build system will 
> produce correct binaries, but the DKMS tool does not differentiate.
> 
> There is a need to ship a different driver source in the DKMS package 
> for RHEL3 as a result (Add a #define SCSI_HAS_VARY_IO in the aacraid.h

> file).

We could do this easily in DKMS with its PATCH_MATCH mechanism, making
it match RHEL kernel names and apply the patch only then which sets
.vary_io=1.

--
Matt Domsch
Sr. Software Engineer, Lead Engineer
Dell Linux Solutions linux.dell.com & www.dell.com/linux Linux on Dell
mailing lists @ http://lists.us.dell.com

^ permalink raw reply	[flat|nested] 6+ messages in thread

* RE: Aacraid - terrible performance on RHES3 pedge 2450
@ 2004-12-08 14:58 Zoe May
  0 siblings, 0 replies; 6+ messages in thread
From: Zoe May @ 2004-12-08 14:58 UTC (permalink / raw)
  To: linux-raid

After trying various Linux distros without success I decided to take the
hardware RAID out of the equation.
If I put the raid card into scsi mode and configure the disks (my system
has 4) as volumes my problem goes away - except that I now can't boot
into Linux. (install completes without issue).

I get as far as the controller picking up the volumes then at the point
when it would normally go into grub I just get a blinking cursor. 
I can see my install by using linux rescue and when I run my disk i/o
tests the results are good -  virtually no CPU wait at all. 
I've tried using one volume with no software RAID, 4 volumes with
software RAID 5, 2 volumes with software RAID 1, all had the same
result.
Tried clearing nvram, flashing the BIOS, flashing the controller. In
desperation removed the RAID rom and ram and set the main BIOS
controller setting to SCSI. The system won't boot (stops just after
processor detection) or let me into 
BIOS when I have this option set. I had to put the RAID hardware back in
just to access the BIOS!

Anyone any suggestions? I can't see any way out of this and it's held up
an upgrade project for weeks now.
At this stage I would be willing to suggest buying another controller if
there is a better one that's 
guaranteed to work with the 2450 and RHEL3. 
Thing I'm worried about is that the next server in the upgrade chain has
exactly the same hardware...

Cheers

Zoe

-----Original Message-----
From: Zoe May 
Sent: Wednesday, November 24, 2004 2:07 PM
To: 'Salyzyn, Mark'
Cc: 'Matt_Domsch@dell.com'
Subject: RE: Aacraid - terrible performance on RHES3 pedge 2450

I tried this but unfortunately no improvement. 
I really need to get this box in to production so I'm going to drag
myself away from AS3 try some different distros. Tried Fedora C2 this
morning but had the same issues. 
Interestingly when I tried using Knoppix I didn't see much improvement
in disk i/o but vmstat didn't output the same high CPU wait times that
seem to be the cause of stopping everything else on the box running as
soon as I start writing to disk.

Thanks for your help

Zoe
-----Original Message-----
From: Salyzyn, Mark [mailto:mark_salyzyn@adaptec.com]
Sent: Thursday, November 18, 2004 4:06 PM
To: Zoe May
Cc: Matt_Domsch@dell.com
Subject: RE: Aacraid - terrible performance on RHES3 pedge 2450

Yes on the directory. No need to rebuild the kernel (if a past build is
sitting there), just do (2.4 kernel):

# make SUBDIRS=drivers/scsi/aacraid modules

and use the replacement module.

Sincerely -- Mark Salyzyn



-----Original Message-----
From: Zoe May [mailto:Zoe.May@Cartesian.co.uk]
Sent: Thursday, November 18, 2004 10:26 AM
To: Salyzyn, Mark
Cc: Matt_Domsch@dell.com
Subject: RE: Aacraid - terrible performance on RHES3 pedge 2450

Hi

After saying I would try the driver attached I opened the attachment and
realised I wasn't sure what to do with the files. Do I need to replace
the files in /usr/src/linux-xxx/drivers/scsi/aacraid with these and
rebuild the kernel?

Thanks

Zoe

-----Original Message-----
From: Salyzyn, Mark [mailto:mark_salyzyn@adaptec.com]
Sent: Monday, November 01, 2004 4:08 PM
To: Zoe May; Matt Domsch
Cc: linux-raid@vger.kernel.org
Subject: RE: Aacraid - terrible performance on RHES3 pedge 2450

Vary_io was added in 1.1.5-2326 to the adaptec aacraid sources, so
probably *none* of these. Adding a ".vary_io = 1," to the
scsi_host_template in the linit.c file would have to be manually added.

I have enclosed the sources of our latest driver that my comments were
based on. This driver has several performance enhancements (depending on
the adapter, between 15% and 45% improvement separate from the .vary_io
RHEL3 enhancement).

Sincerely -- Mark Salyzyn

-----Original Message-----
From: Zoe May [mailto:Zoe.May@Cartesian.co.uk]
Sent: Monday, November 01, 2004 10:52 AM
To: Matt Domsch; Salyzyn, Mark
Cc: linux-raid@vger.kernel.org
Subject: RE: Aacraid - terrible performance on RHES3 pedge 2450

Thanks for the info

I've upgraded the kernel to 2.4.21-20 and have a very slight increase in
performance but unfortunately not much to write home about.
I have limited experience with kernel compilation and I'm a little
confused as to which aacraid.h I would have to add the #define
SCSI_HAS_VARY_IO 1 to.
 
I have the following aacraid.h files

/usr/src/aacraid-1.1.4-2302/aacraid.h
/usr/src/linux-2.4.21-20.EL/drivers/addon/aacraid_10102/aacraid.h
/usr/src/linux-2.4.21-20.EL/drivers/scsi/aacraid/aacraid.h

Ordinarily I would modify the .h file in the scsi/aacraid dir and then
go through the full kernel recompile but I have updated aacraid using
the Dell rpm. I assume the version in this folder is the old version
that shipped with the 2.4.21-20 source. If there is a way that I can
recompile just the aacraid module without doing a full kernel recompile
I would prefer to do this .


Zoe


-----Original Message-----
From: Matt Domsch [mailto:Matt_Domsch@dell.com]
Sent: Monday, October 25, 2004 2:02 PM
To: Salyzyn, Mark
Cc: Zoe May; linux-raid@vger.kernel.org
Subject: Re: Aacraid - terrible performance on RHES3 pedge 2450

On Mon, Oct 25, 2004 at 08:28:47AM -0400, Salyzyn, Mark wrote:
> Vary_io may be the reason for this.
> 
> The driver needs to have .vary_io set to 1.
> 
> There are no kernel clues to its existence (that I know of), but the 
> driver has an SCSI_HAS_VARY_IO to enable. Our build system will 
> produce correct binaries, but the DKMS tool does not differentiate.
> 
> There is a need to ship a different driver source in the DKMS package 
> for RHEL3 as a result (Add a #define SCSI_HAS_VARY_IO in the aacraid.h

> file).

We could do this easily in DKMS with its PATCH_MATCH mechanism, making
it match RHEL kernel names and apply the patch only then which sets
.vary_io=1.

--
Matt Domsch
Sr. Software Engineer, Lead Engineer
Dell Linux Solutions linux.dell.com & www.dell.com/linux Linux on Dell
mailing lists @ http://lists.us.dell.com

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2004-12-08 14:58 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-10-25 12:28 Aacraid - terrible performance on RHES3 pedge 2450 Salyzyn, Mark
2004-10-25 13:01 ` Matt Domsch
  -- strict thread matches above, loose matches on Subject: below --
2004-12-08 14:58 Zoe May
2004-11-01 15:51 Zoe May
2004-10-25 10:31 Zoe May
2004-10-25 12:22 ` Matt Domsch

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.